Optimising ASReview Simulations: A Generic Multiprocessing Solution for 'light-data' and 'heavy-data' Users

Abstract

Active learning (AL) aided systematic review pipelines are a promising tool for optimising and speeding up the performance of systematic reviews. In this paper, we present an architecture design with multiprocessing computational strategy for ASReview Makita workflow generator (Teijema, Van de Schoot et al., 2023) using Kubernetes software for the purposes of deployment with cloud technologies. The main goal of this study is to contribute to the following research in the study field of AL-assisted systematic review simulations by focusing on the parallel and distributed computing techniques. We provide an in detail technical explanation of the proposed cloud architecture and its usage manual. In addition to that, we conducted 1140 simulations studies investigating computational time required for ARFI Makita template using various number of CPUs and RAM settings. Our analysis demonstrates the degree to which ARFI Template can be accelerated with multiprocessing computing usage. The parallel computation strategy and the architecture design which were developed in the present paper can contribute to future research with more optimal simulation time and, at the same time, ensure the safe completion of the needed processes.